Bacterial cells contain three types of DNA polymerases termed polymerase I, II and III. DNA polymerase III (pol III) is responsible for the replication of the majority of the chromosome. Pol III is referred to as a replicative polymerase; replicative polymerases are rapid and highly processive enzymes. Pol I and II are referred to as non-replicative polymerases although both enzymes appear to have roles in replication. DNA polymerase I is the most abundant polymerase and is responsible for some types of DNA repair, including a repair-like reaction that permits the joining of Okazaki fragments during DNA replication. Pol I is essential for the repair of DNA damage induced by WV irradiation and radiomimetic drugs. Pol II is thought to play a role in repairing DNA damage which induces the SOS response and in mutants which lack both pol I and III, pol II repairs UV-induced lesions. Pol I and II are monomeric polymerases while pol III comprises a multisubunit complex.
In E. coli, pol III comprises the catalytic core of the E. coli replicase. In E. coli, there are approximately 400 copies of DNA polymerase I per cell, but only 10-20 copies of Pol III (Kornberg and Baker, DNA Replication, 2d ed., W.H. Freeman & Company, [1992], pp. 167; and Wu et al. J. Biol. Chem., 259:12117-12122 [1984]). The low abundance of Pol III and its relatively feeble activity on gapped DNA templates typically used as a general replication assays delayed its discovery until the availability of mutants defective in DNA polymerase I (Kornberg and Gefter, J. Biol. Chem., 47:5369-5375 [1972]).
The catalytic subunit of Pol III is distinguished as a component of E. coli major replicative complex, apparently not by its intrinsic catalytic activity, but by its ability to interact with other replication proteins at the fork. These interactions confer upon the enzyme enormous processivity. Once the DNA polymerase III holoenzyme associates with primed DNA, it does not dissociate for over 40 minutes-the time required for the synthesis of the entire 4 Mb E. coli chromosome (McHenry, Ann. Rev. Biochem., 57:519-550 [1988]). Studies in coupled rolling circle models of the replication fork suggest the enzyme can synthesize DNA 150 kb or longer without dissociation in vitro (Mok and Marians, J. Biol. Chem., 262:16644-16654 [1987]; Wu et al., J. Biol. Chem., 267:4030-4044 [1992]). The essential interaction required for this high processivity is an interaction between the a catalytic subunit and a dimer of A, a sliding clamp processivity factor that encircles the DNA template like a bracelet, permitting it to rapidly slide along with the associated polymerase, but preventing it from falling off (LaDuca et al., J. Biol. Chem., 261:7550-7557 [1986]; Kong et al., Cell 69:425-437 [1992]). The .beta.-.alpha. association apparently retains the polymerase on the template during transient thermal fluctuations when it might otherwise dissociate.
The .beta.2 bracelet cannot spontaneously associate with high molecular weight DNA, it requires a multiprotein DnaX-complex to open and close it around DNA using the energy of ATP hydrolysis (Wickner, Proc. Natl. Acad. Sci. USA 73:35411-3515 [1976]; Naktinis et al., J. Biol. Chem., 270:13358-13365 [1985]; and Dallmann et al., J. Biol. Chem., 270:29555-29562 [1995]). In E. coli, the dnaX gene encodes two proteins, .tau. and .gamma.. .gamma. is generated by a programmed ribosomal frameshifting mechanism five-sevenths of the way through dnaX MRNA, placing the ribosome in a -1 reading frame where it immediately encounters a stop codon (Flower and McHenry Proc. Natl. Acad. Sci. USA 87:3713-3717 [1990]; Blinkowa and Walker, Nucl. Acids Res., 18:1725-1729 [1990]; and Tsuchihashi and Kornberg, Proc. Natl. Acad. Sci. USA 87:2516-2520 [1990]). In E. coli, the DnaX-complex has the stoichiometry .gamma..sub.2.tau..sub.2.delta..sub.1.delta.'.sub.1.chi..sub.1.iota..sub.1 (Dallmann and McHenry, J. Biol. Chem., 270:29563-29569[1995]). The .tau. protein contains an additional carboxyl-terminal domain that interacts tightly with the polymerase, holding two polymerases together in one complex that can coordinately replicate the leading and lagging strand of the replication fork simultaneously (McHenry, J. Biol. Chem., 257:2657-2663 [1982]; Studwell and O'Donnell, Biol. Chem., 266:19833-19841 [1991]; McHenry, Ann. Rev. of Biochem. 57:519-550 [1988]).
Pol IIIs are apparently conserved throughout mesophilic eubacteria. In addition to E. coli and related proteobacteria, the enzyme has been purified from the firmicute Bacillus subtilis (ow et al., J. Biol. Chem., 251:1311-1325 [1976]; Hammond and Brown [1992]). With the proliferation of bacterial genomes sequenced, by inference from DNA sequence, pol III exits in organisms as widely divergent as Caulobacter, Mycobacteria, Mycoplasma, B. subtilis and Synechocystis. The existence of dnaX and dnaN (structural gene for .beta.) is also apparent in these organisms. These general replication mechanisms are conserved even more broadly in biology. Although eukaryotes do not contain polymerases homologous to Pol III, eukaryotes contain special polymerases devoted to chromosomal replication and .beta.-like processivity factors (PCNA) and DnaX-like ATPases (RFC, Activator I) that assemble these processivity factors on DNA (Yoder and Burgers, J. Biol. Chem., 266:22689-22697 [1991]; Brush and Stillman, Meth. Enzymol., 262:522-548 [1995]; Uhlmann et al., Proc. Natl. Acad. Sci. USA 93:6521-6526 [1996]).
In spite of the apparent ubiquity of Pol Ills and their associated factors required to function as a replicase, the identification of such enzymes remains to be accomplished for many other organisms.